ACS Catalysis
Letter
3
a
a
Scheme 3. Scope of Aliphatic C(sp )−H Coupling Partners
Scheme 4. Scope of Allylating Agents
a
Reaction conditions: 1 (0.1 mmol), 2 (2 mmol), allylating agent (0.2
mmol), Pd (dba) (2.5 mol %), P(3,5-Me C H ) (6 mol %), PT (5
2
3
2
6
3 3
mol %), DCE (1 mL), 6 W LEDs (λ = 420 nm), under N , room
max
2
b
c
temperature, 36 h. Run for 48 h. 5:1 E/Z. VEC, vinyl ethylene
carbonate. EB, 1,2-epoxybutene.
Furthermore, vinyl ethylene carbonate and 1,2-epoxybutene
were also found to be applicable in this protocol to yield the
same allylation product containing an allylic alcohol moiety,
thus providing a useful handle for further functionalization.
To understand the reaction mechanism, a series of
mechanistic studies were performed. A radical trapping
experiment showed that the addition of TEMPO to the
model reaction completely inhibited the formation of the
desired product (Scheme 5A). The Stern−Volmer quenching
experiments revealed that only C−H coupling partners, such as
toluene and cyclohexane, were capable of quenching the
excited state of PT (Scheme 5B). In addition, kinetic studies
established a first-order dependence of the initial rate on the
amount of both the palladium catalyst and HAT photocatalyst
(Scheme 5C), indicating that these two catalysts might
a
Reaction conditions: 1 (0.1 mmol), hydrocarbon (2 mmol), 3 (0.2
mmol), Pd (dba) (2.5 mol %), P(3,5-Me C H ) (6 mol %), PT (5
mol %), DCE (1 mL), 6 W LEDs (λ = 420 nm), under N , room
temperature, 24 h. Run for 36 h. 7:1 r/r. Run for 48 h. Run for 96
h. 1.3:1 dr. 2.3:1 dr.
2
3
2
6
3 3
max
2
b
c
d
e
f
g
2
4) in high yields. Interestingly, adamantane underwent
functionalization predominantly at the sterically hindered
bridgehead position with 7:1 site selectivity (25), presumably
due to the relatively low bond dissociation energies (BDEs) of
3
the tertiary C(sp )−H bonds. Similarly, 2,3-dimethylbutane
and 5-methylhexan-2-one were functionalized at the tertiary
position with high levels of site selectivity (26 and 27).
Toluene and methylarenes bearing various synthetic valuable
functional groups, such as ether, borate, ketone, and ester, all
proceeded smoothly to provide the corresponding products
20,13a,c
function in a synergistic manner
and be relevant to the
rate-determining step of this binary catalysis. Moreover, the
treatment of a potential Giese addition product 50 and methyl
allyl carbonate 3 with 2.5 mol % Pd (dba) and 6 mol % P(3,5-
(28−32). Notably, 2-methylthiophene could be converted to
2
3
the corresponding product (33) in 42% yield. Moreover,
ethylbenzene, 2-propylbenzene, and cyclohexylbenzene also
underwent site-selective benzylic C−H functionalization to
provide the corresponding products (34−36) in moderate
yields, albeit with 1.3:1 diastereoselectivity for ethylbenzene. In
addition, methyl tert-butyl ether, 1,4-dioxane, and 1,3-
benzodioxole were viable to afford the α-oxy functionalized
products (37−39) in good yields. Intriguingly, cyclopentyl
methyl ether was specifically functionalized at the tertiary
position adjacent to the oxygen, leading to a sterically hindered
ether in good yield (40). It was noteworthy that N-
methyldiphenylamine and tert-butyl dimethylcarbamate were
suitable for the reaction to generate the desired products (41
and 42) in good yields.
Me C H ) was unable to generate the desired allylation
2 6 3 3
product 4, and nearly complete starting material 50 was
recovered (Scheme 5D). In addition, the Giese addition
product 50 was not observed in the progress curve under the
optimized conditions (Scheme 5D). These results suggested
that the Giese addition product 50 had extremely low reactivity
under the optimized condition; therefore, a sequential reaction
of Giese addition/Pd-catalyzed allylic alkylation by using a
Giese addition product as a key intermediate was unfeasible.
Notably, a series of control experiments using hexafluoroiso-
propanol (HFIP) as an additive showed that the increasing
amount of HFIP as an additive could significantly inhibit the
formation of the desired allylation product 4; meanwhile, the
generation of the Giese addition product 50 was gradually
increased (Scheme 5D). In particular, the presence of 60 mol
% HFIP resulted in the complete formation of the Giese
addition product 50. These results were in accordance with a
two-electron allylation pathway, where a carbanionic species
served as a key intermediate for nucleophile attack on the π-
allylpalladium intermediate; therefore, the presence of HFIP as
As for the allylating agents, a range of substituted allyl
carbonates were proven to be competent for this protocol
(
carbonates were successfully employed to provide the
corresponding products (43−47) in moderate to good yields.
Interestingly, the reaction with 1-phenylallyl carbonate resulted
in the generation of the allylation product (48) as a 5:1 E/Z
isomeric mixture, in which the thermodynamically less stable Z
Scheme 4). Both 2-alkyl- and 2-aryl-substituted allyl
21
a proton-transfer shuttle would enhance the competing
protonation to provide the Giese addition product. On the
contrary, a single-electron allylation pathway was conducted
19
isomer might be raised from a photochemical isomerization.
6
759
ACS Catal. 2021, 11, 6757−6762
Shen, Yang
